Nondaily Smokers’ Changes in Cigarette Consumption With Very Low-Nicotine-Content Cigarettes: A Randomized Double-blind Clinical Trial | Lifestyle Behaviors | JAMA Psychiatry | JAMA Network
[Skip to Navigation]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Navigation Landing]
Figure 1.  Study Flowchart
Study Flowchart

CO indicates carbon monoxide; RYO, roll your own.

Figure 2.  Mean Cigarettes per Day (CPD) at Baseline and End of Study
Mean Cigarettes per Day (CPD) at Baseline and End of Study

Baseline was 2 weeks before randomization, and weeks 9 to 10 was the end of the study. Observed data are plotted (mean [SE]). Baseline data (own-brand cigarettes provided gratis) are based on observed data, and data for weeks 9 to 10 are based either on observed data (participants not lost to observation) or monotone multiple imputation of missing CPD. Analysis of imputed data indicated that, while controlling for baseline CPD, the very low-nicotine-content cigarettes (VLNCC) group reduced their CPD significantly more than the normal-nicotine-content cigarettes (NNCC) group during the 10-week postrandomization period (P < .001). Mean decrease among VLNCC participants was 1.51 CPD (95% CI, 0.86-2.17) more than NNCC participants. Analysis of observed data showed substantively the same effect (P < .001). Mean decrease among VLNCC participants was 1.64 CPD (95% CI, 0.93-2.35) more than NNCC participants.

Figure 3.  Percentage of Days Smoked at Baseline vs End of Study
Percentage of Days Smoked at Baseline vs End of Study

Data are mean percentages of days with any smoking during the baseline and weeks 9 to 10 by treatment group. Observed data are plotted (mean [SE]). Analysis indicated that, while controlling for baseline smoking day percentage, the very low-nicotine-content cigarettes (VLNCC) group (n = 91) reduced their days smoking percentage significantly more than the normal-nicotine-content cigarettes (NNCC) group (n = 99) over the 10-week postrandomization period (P < .001). Mean decrease among the VLNCC group was 17% (95% CI, 9%-25%) more than the NNCC group.

Figure 4.  Cigarette Consumption by Time and Treatment Group
Cigarette Consumption by Time and Treatment Group

Mean daily cigarette consumption (observed data are plotted as mean [SE]) in 2-week periods among participants in the very low-nicotine-content cigarettes (VLNCC) and normal-nicotine-content cigarettes (NNCC) groups. Both linear and quadratic trends in cigarettes per day (CPD) varied significantly by group.

Table.  Participant Characteristics
Participant Characteristics
Supplement 2.

eTable 1. Inclusion and exclusion criteria

eTable 2. Missing data patterns

eTable 3. Cigarettes per day (CPD) and primary outcome model imputation variance information

eFigure 1. Study schema

eFigure 2. Log-transformed cigarettes smoked per day at baseline vs end of study

eFigure 3. Cigarettes smoked per day on days smoked, at baseline vs end of study

eFigure 4. Cigarette consumption, at baseline vs end of study, for male and female subjects

eFigure 5. Cigarette consumption, at baseline and end of study, for African American and White Subjects

eFigure 6. Cigarette consumption, at baseline vs end of study, for CITS and NITS

eFigure 7. Cigarette consumption, at baseline vs end of study, for subjects stratified by baseline CPD

eFigure 8. Cigarette consumption, at baseline vs end of study, for subjects stratified by baseline number of days per week smoking

eFigure 9. Cigarette consumption at baseline vs end of study among subjects with low self-reported smoking conventional cigarettes

eFigure 10. Cigarettes smoked per day (Research Cigarettes Only), at baseline (2 weeks prior to randomization) vs end of study (weeks 9-10)

eFigure 11. Cigarette consumption at baseline vs end of study, excluding subjects inferred to have cheated with conventional cigarettes, based on urinary cotinine values

eFigure 12. Mean total urinary cotinine concentrations, at baseline and end of study

eFigure 13. Urinary total cotinine concentrations by time and treatment group

eFigure 14. Reported new use of e-cigarettes post-randomization, by treatment group and baseline smoking rate

eFigure 15. Cigarette consumption at baseline vs end of study among subjects who did not increase their usage of other nicotine-containing products vs baseline

1.
Jacobs  EJ, Newton  CC, Carter  BD,  et al.  What proportion of cancer deaths in the contemporary United States is attributable to cigarette smoking?  Ann Epidemiol. 2015;25(3):179-182.e1.PubMedGoogle ScholarCrossref
2.
Benowitz  NL.  Nicotine addiction.  N Engl J Med. 2010;362(24):2295-2303. doi:10.1056/NEJMra0809890PubMedGoogle ScholarCrossref
3.
Stolerman  IP, Jarvis  MJ.  The scientific case that nicotine is addictive.  Psychopharmacology (Berl). 1995;117(1):2-10. doi:10.1007/BF02245088PubMedGoogle ScholarCrossref
4.
Benowitz  NL, Henningfield  JE.  Establishing a nicotine threshold for addiction: the implications for tobacco regulation.  N Engl J Med. 1994;331(2):123-125. doi:10.1056/NEJM199407143310212PubMedGoogle ScholarCrossref
5.
Henningfield  JE, Benowitz  NL, Slade  J, Houston  TP, Davis  RM, Deitchman  SD.  Reducing the addictiveness of cigarettes: Council on Scientific Affairs, American Medical Association.  Tob Control. 1998;7(3):281-293. doi:10.1136/tc.7.3.281PubMedGoogle ScholarCrossref
6.
Family Smoking Prevention and Tobacco Control Act and Federal Retirement Reform, HR 1256, 111th Cong, 1st Sess (PA 1995).
7.
Tobacco product standard for nicotine level of combusted cigarettes. Fed Regist. 2018;83:11818-11843. To be codified at 21 CFR §1130.
8.
FDA announces comprehensive regulatory plan to shift trajectory of tobacco-related disease, death [press release]. Silver Spring, MD: US Food and Drug Administration; July 28, 2017.
9.
Benowitz  NL, Hall  SM, Stewart  S, Wilson  M, Dempsey  D, Jacob  P  III.  Nicotine and carcinogen exposure with smoking of progressively reduced nicotine content cigarette.  Cancer Epidemiol Biomarkers Prev. 2007;16(11):2479-2485. doi:10.1158/1055-9965.EPI-07-0393PubMedGoogle ScholarCrossref
10.
Donny  EC, Houtsmuller  E, Stitzer  ML.  Smoking in the absence of nicotine: behavioral, subjective and physiological effects over 11 days.  Addiction. 2007;102(2):324-334. doi:10.1111/j.1360-0443.2006.01670.xPubMedGoogle ScholarCrossref
11.
Donny  EC, Denlinger  RL, Tidey  JW,  et al.  Randomized trial of reduced-nicotine standards for cigarettes.  N Engl J Med. 2015;373(14):1340-1349. doi:10.1056/NEJMsa1502403PubMedGoogle ScholarCrossref
12.
Jamal  A, King  BA, Neff  LJ, Whitmill  J, Babb  SD, Graffunder  CM.  Current cigarette smoking among adults: United States, 2005-2015.  MMWR Morb Mortal Wkly Rep. 2016;65(44):1205-1211. doi:10.15585/mmwr.mm6544a2PubMedGoogle ScholarCrossref
13.
US Department of Health and Human Services.  Results from the 2013 national survey on drug use and health: summary of national findings. https://www.samhsa.gov/data/sites/default/files/NSDUHresultsPDFWHTML2013/Web/NSDUHresults2013.pdf. Accessed June 5, 2018.
14.
Reyes-Guzman  CM, Pfeiffer  RM, Lubin  J,  et al.  Determinants of light and intermittent smoking in the United States: results from three pooled national health surveys.  Cancer Epidemiol Biomarkers Prev. 2017;26(2):228-239. doi:10.1158/1055-9965.EPI-16-0028PubMedGoogle ScholarCrossref
15.
Inoue-Choi  M, Liao  LM, Reyes-Guzman  C, Hartge  P, Caporaso  N, Freedman  ND.  Association of long-term, low-intensity smoking with all-cause and cause-specific mortality in the National Institutes of Health-AARP Diet and Health Study.  JAMA Intern Med. 2017;177(1):87-95. doi:10.1001/jamainternmed.2016.7511PubMedGoogle ScholarCrossref
16.
Schane  RE, Ling  PM, Glantz  SA.  Health effects of light and intermittent smoking: a review.  Circulation. 2010;121(13):1518-1522. doi:10.1161/CIRCULATIONAHA.109.904235PubMedGoogle ScholarCrossref
17.
Tindle  HA, Shiffman  S.  Smoking cessation behavior among intermittent smokers versus daily smokers.  Am J Public Health. 2011;101(7):e1-e3. doi:10.2105/AJPH.2011.300186PubMedGoogle ScholarCrossref
18.
Shiffman  S, Dunbar  MS, Benowitz  NL.  A comparison of nicotine biomarkers and smoking patterns in daily and nondaily smokers.  Cancer Epidemiol Biomarkers Prev. 2014;23(7):1264-1272. doi:10.1158/1055-9965.EPI-13-1014PubMedGoogle ScholarCrossref
19.
Shiffman  S, Dunbar  MS, Tindle  HA, Ferguson  SG.  Nondaily smokers’ experience of craving on days they do not smoke.  J Abnorm Psychol. 2015;124(3):648-659. doi:10.1037/abn0000063PubMedGoogle ScholarCrossref
20.
Shiffman  S, Paty  JA, Gnys  M, Kassel  JD, Elash  C.  Nicotine withdrawal in chippers and regular smokers: subjective and cognitive effects.  Health Psychol. 1995;14(4):301-309. doi:10.1037/0278-6133.14.4.301PubMedGoogle ScholarCrossref
21.
Russell  MA.  Cigarette smoking: natural history of a dependence disorder.  Br J Med Psychol. 1971;44(1):1-16. doi:10.1111/j.2044-8341.1971.tb02141.xPubMedGoogle ScholarCrossref
22.
Shiffman  S, Dunbar  MS, Li  X,  et al.  Smoking patterns and stimulus control in intermittent and daily smokers.  PLoS One. 2014;9(3):e89911. doi:10.1371/journal.pone.0089911PubMedGoogle ScholarCrossref
23.
Shiffman  S, Tindle  H, Li  X, Scholl  S, Dunbar  M, Mitchell-Miland  C.  Characteristics and smoking patterns of intermittent smokers.  Exp Clin Psychopharmacol. 2012;20(4):264-277. doi:10.1037/a0027546PubMedGoogle ScholarCrossref
24.
Shiffman  S, Dunbar  MS, Kirchner  TR,  et al.  Cue reactivity in converted and native intermittent smokers.  Nicotine Tob Res. 2015;17(1):119-123. doi:10.1093/ntr/ntu147PubMedGoogle ScholarCrossref
25.
Trinidad  DR, Pérez-Stable  EJ, Emery  SL, White  MM, Grana  RA, Messer  KS.  Intermittent and light daily smoking across racial/ethnic groups in the United States.  Nicotine Tob Res. 2009;11(2):203-210. doi:10.1093/ntr/ntn018PubMedGoogle ScholarCrossref
26.
Perkins  KA, Jacobs  L, Sanders  M, Caggiula  AR.  Sex differences in the subjective and reinforcing effects of cigarette nicotine dose.  Psychopharmacology (Berl). 2002;163(2):194-201. doi:10.1007/s00213-002-1168-1PubMedGoogle ScholarCrossref
27.
Shiffman  S, Scholl  S.  Increases in cigarette consumption and decreases in smoking intensity when non-daily smokers are provided with free cigarettes.  Nicotine Tob Res. 2017; Advance online publication. doi:10.1093/ntr/ntx221PubMedGoogle Scholar
28.
Twenty Second Century. SPECTRUM Government Research Cigarettes. http://www.xxiicentury.com/.
29.
Blank  MD, Breland  AB, Enlow  PT, Duncan  C, Metzger  A, Cobb  CO.  Measurement of smoking behavior: comparison of self-reports, returned cigarette butts, and toxicant levels.  Exp Clin Psychopharmacol. 2016;24(5):348-355. doi:10.1037/pha0000083PubMedGoogle ScholarCrossref
30.
Sobell  LC, Sobell  MB, Maisto  SA. Time-line follow-back assessment methods. In: Lettieri  DJ, Sayers  MA, Nelson  JE, eds.  NIAA Treatment Handbook Series: Vol. 2, Alcoholism Treatment Assessment Research Instruments. Washington, DC: National Institute on Alcoholism and Drug Abuse DHHS Publications; 1985:530-534.
31.
Wray  JM, Gass  JC, Miller  EI, Wilkins  DG, Rollins  DE, Tiffany  ST.  A comparative evaluation of self-report and biological measures of cigarette use in nondaily smokers.  Psychol Assess. 2016;28(9):1043-1050. doi:10.1037/pas0000227PubMedGoogle ScholarCrossref
32.
Shiffman  S, Scholl  SM.  Three approaches to quantifying cigarette consumption: data from nondaily smokers.  Psychol Addict Behav. 2018;32(2):249-254. doi:10.1037/adb0000346PubMedGoogle ScholarCrossref
33.
SRNT Subcommittee on Biochemical Verification.  Biochemical verification of tobacco use and cessation.  Nicotine Tob Res. 2002;4(2):149-159. doi:10.1080/14622200210123581PubMedGoogle ScholarCrossref
34.
Hughes  JR, Keely  JP, Niaura  RS, Ossip-Klein  DJ, Richmond  RL, Swan  GE.  Measures of abstinence in clinical trials: issues and recommendations.  Nicotine Tob Res. 2003;5(1):13-25. doi:10.1080/1462220031000070552PubMedGoogle ScholarCrossref
35.
West  R, Hajek  P, Stead  L, Stapleton  J.  Outcome criteria in smoking cessation trials: proposal for a common standard.  Addiction. 2005;100(3):299-303. doi:10.1111/j.1360-0443.2004.00995.xPubMedGoogle ScholarCrossref
36.
Murphy  SE, Park  SS, Thompson  EF,  et al.  Nicotine N-glucuronidation relative to N-oxidation and C-oxidation and UGT2B10 genotype in five ethnic/racial groups.  Carcinogenesis. 2014;35(11):2526-2533. doi:10.1093/carcin/bgu191PubMedGoogle ScholarCrossref
37.
Pitt  JJ.  Principles and applications of liquid chromatography-mass spectrometry in clinical biochemistry.  Clin Biochem Rev. 2009;30(1):19-34.PubMedGoogle Scholar
38.
Nardone  N, Donny  EC, Hatsukami  DK,  et al.  Estimations and predictors of non-compliance in switchers to reduced nicotine content cigarettes.  Addiction. 2016;111(12):2208-2216. doi:10.1111/add.13519PubMedGoogle ScholarCrossref
39.
Benowitz  NL, Nardone  N, Hatsukami  DK, Donny  EC.  Biochemical estimation of noncompliance with smoking of very low nicotine content cigarettes.  Cancer Epidemiol Biomarkers Prev. 2015;24(2):331-335. doi:10.1158/1055-9965.EPI-14-1040PubMedGoogle ScholarCrossref
40.
Wang  B, Fang  Y, Jin  M.  Combining type-III analyses from multiple imputations. http://support.sas.com/resources/papers/proceedings14/1543-2014.pdf. Accessed June 5, 2018.
41.
Hatsukami  DK, Heishman  SJ, Vogel  RI,  et al.  Dose-response effects of spectrum research cigarettes.  Nicotine Tob Res. 2013;15(6):1113-1121. doi:10.1093/ntr/nts247PubMedGoogle ScholarCrossref
42.
Shiffman  S, Dunbar  MS, Scholl  SM, Tindle  HA.  Smoking motives of daily and non-daily smokers: a profile analysis.  Drug Alcohol Depend. 2012;126(3):362-368. doi:10.1016/j.drugalcdep.2012.05.037PubMedGoogle ScholarCrossref
43.
Singer  JM, Rocha  FMM, Nobre  JS.  Graphical tools for detecting departures from linear mixed model assumptions and some remedial measures.  Int Stat Rev. 2017;85(2):290-324. doi:10.1111/insr.12178Google ScholarCrossref
44.
Edwards  AL.  Multiple Regression and the Analysis of Variance and Covariance New York. NY: Freeman; 1985.
45.
 Heavy: Robust Estimation Using Heavy-Tailed Distributions. CRAN. https://cran.r-project.org/web/packages/heavy/index.html. Accessed June 5, 2018,
46.
Hatsukami  DK, Kotlyar  M, Hertsgaard  LA,  et al.  Reduced nicotine content cigarettes: effects on toxicant exposure, dependence and cessation.  Addiction. 2010;105(2):343-355. doi:10.1111/j.1360-0443.2009.02780.xPubMedGoogle ScholarCrossref
47.
Cohen  J.  Statistical power analysis for the behavioral sciences. Hillsdale, NJ: Lawrence Erlbaum Associates; 1987.
48.
Rose  JE.  Nicotine and nonnicotine factors in cigarette addiction.  Psychopharmacology (Berl). 2006;184(3-4):274-285. doi:10.1007/s00213-005-0250-xPubMedGoogle ScholarCrossref
49.
Schane  RE, Glantz  SA, Ling  PM.  Nondaily and social smoking: an increasingly prevalent pattern.  Arch Intern Med. 2009;169(19):1742-1744. doi:10.1001/archinternmed.2009.315PubMedGoogle ScholarCrossref
50.
Shiffman  S, Dunbar  MS, Ferguson  SG.  Stimulus control in intermittent and daily smokers.  Psychol Addict Behav. 2015;29(4):847-855. doi:10.1037/adb0000052PubMedGoogle ScholarCrossref
51.
Farsalinos  KE, Polosa  R.  Safety evaluation and risk assessment of electronic cigarettes as tobacco cigarette substitutes: a systematic review.  Ther Adv Drug Saf. 2014;5(2):67-86. doi:10.1177/2042098614524430PubMedGoogle ScholarCrossref
52.
Polosa  R, Rodu  B, Caponnetto  P, Maglia  M, Raciti  C.  A fresh look at tobacco harm reduction: the case for the electronic cigarette.  Harm Reduct J. 2013;10:19. doi:10.1186/1477-7517-10-19PubMedGoogle ScholarCrossref
53.
Royal College of Physicians.  Nicotine without smoke: tobacco harm reduction. https://www.rcplondon.ac.uk/projects/outputs/nicotine-without-smoke-tobacco-harm-reduction-0. Accessed June 5, 2018.
54.
Apelberg  BJ, Feirman  SP, Salazar  E,  et al.  Potential public health effects of reducing nicotine levels in cigarettes in the United States.  N Engl J Med. 2018;378(18):1725-1733. doi:10.1056/NEJMsr1714617PubMedGoogle ScholarCrossref
Original Investigation
October 2018

Nondaily Smokers’ Changes in Cigarette Consumption With Very Low-Nicotine-Content Cigarettes: A Randomized Double-blind Clinical Trial

Author Affiliations
  • 1Department of Psychology, University of Pittsburgh, Pittsburgh, Pennsylvania
  • 2UPMC Hillman Cancer Center, Pittsburgh, Pennsylvania
  • 3Department of Biostatistics, University of Pittsburgh, Pittsburgh, Pennsylvania
JAMA Psychiatry. 2018;75(10):995-1002. doi:10.1001/jamapsychiatry.2018.1831
Key Points

Question  Do nondaily smokers decrease cigarette consumption when switched from their own brand to very low-nicotine-content cigarettes (VLNCC) compared with identical normal-nicotine-content cigarettes (NNCC)?

Findings  In this randomized double-blind clinical trial with 238 randomized participants, the mean cigarettes per day was reduced significantly more in the VLNCC group (1.6 cigarettes per day, 51% of baseline) than in the NNCC group (0.05 cigarettes per day, 2% of baseline).

Meaning  Like daily smokers, switching to VLNCCs causes nondaily smokers to reduce their cigarette consumption, although it does not necessarily cause them to stop smoking.

Abstract

Importance  The US Food and Drug Administration is considering limiting cigarettes to very low nicotine levels. Cigarette consumption of nondaily intermittent smokers (ITS), who compose one-third of US adult smokers, could feasibly increase or could be unaffected if their smoking is not motivated by nicotine seeking.

Objective  To compare cigarette consumption in ITS receiving very low-nicotine-content cigarettes (VLNCCs) or identical normal-nicotine-content cigarettes (NNCCs).

Design, Setting, and Participants  This randomized double-blind clinical trial was conducted from June 2015 to July 2017 at a single US site. Volunteer ITS not planning to quit were recruited via media. Overall, 297 individuals enrolled, and 238 were randomized. Analyses were intent-to-treat.

Interventions  After a 2-week baseline of smoking their own brand of cigarettes provided gratis, ITS were randomized to VLNCCs or NNCCs for 10 weeks.

Main Outcomes and Measures  The number of cigarettes per day (CPD) was assessed by real-time reporting, timeline follow-back reports, and cigarette butt counts. The primary outcome was change in CPD from baseline to weeks 9 to 10 of intervention, adjusting for baseline CPD.

Results  The mean (SD) age of the 238 randomized participants was 37.9 (13.8) years. Of 238 participants, 108 (45%) were men. At baseline, the mean (SD) CPD was 3.1 (2.9). In intent-to-treat analyses using multiple imputation to address missing data, the VLNCC group had a mean decrease of 1.6 CPD (95% CI, 1.1-2.0; 51% of baseline) vs 0.05 decrease with NNCCs (95% CI, −0.5 to 0.4; 2% of baseline). Treatment group differences were not materially moderated by sex, race/ethnicity, or history of daily smoking. Cheating with conventional cigarettes, inferred from cotinine assays, was more common in the VLNCC group (OR, 2.95; 95% CI, 1.54-5.66), but sensitivity analyses showed significant VLNCC effects among the compliant participants as well. In longitudinal analysis of CPD over time with random intercept and slope, the VLNCC and NNCC groups differed significantly in both linear (−0.15; 95% CI, −0.22 to −0.08; P < .001) and quadratic (0.0026; 95% CI, 0.0010-0.0042; P = .002) trends: CPD dropped by 43.8% in the VLNCC group over 4 weeks, then leveled off thereafter. Abstinence (intent-to-treat, biochemically verified) in weeks 9 to 10 postrandomization did not differ significantly by treatment group (VLNCC, 10.2% vs NNNC, 5.0%; P = .28).

Conclusions and Relevance  Switching to VLNCCs caused substantial smoking reduction among ITS but did not significantly increase abstinence. Response to a VLNCC intervention suggests that nicotine-seeking motivates ITS’ smoking.

Trial Registration  ClinicalTrials.gov Identifier: NCT02228824

Introduction

Tobacco smoking, the leading cause of preventable mortality,1 is typically maintained by nicotine dependence.2,3 This underlies a policy proposal that smoking could be reduced or eliminated if the nicotine levels in tobacco were reduced to a level too low to initiate or maintain dependence.4,5 The US Food and Drug Administration, which regulates tobacco,6 announced in 2018 that it is considering mandating such reductions.7,8

Consistent with several small studies,9,10 a large study of daily smokers (≥10 cigarettes per day) by Donny et al11 reported that very low-nicotine-content cigarettes (VLNCCs) reduced mean consumption by about 5 cigarettes a day, a decrease of 23% to 30%.

However, 25% to 33% of adult US smokers do not smoke daily.12-14 These intermittent smokers (ITS) nevertheless experience health risks from smoking15,16 and have low quit rates.17 It is not clear how ITS might respond to VLNCCs. It is not established whether ITS’ smoking is motivated by nicotine-seeking. Intermittent smokers absorb normal amounts of nicotine18 but do not experience craving or withdrawal when abstaining.19,20 If ITS smoke for the acute effects of nicotine,21 smoking VLNCCs might cause them to increase smoking. While heavy daily smokers might find it challenging to substantially increase their cigarette consumption, ITS might find it more feasible, especially on occasions when they particularly seek nicotine.22 Thus, it is important to evaluate the effect of VLNCCs on ITS’ smoking, both to assess the impact of a universal VLNCC policy and to address the role of nicotine in motivating ITS’ smoking.

Several participant characteristics might moderate ITS’ responses to VLNCCs. Some ITS have a history of daily smoking,23 and these converted ITS show greater nicotine dependence24 and may respond more like daily smokers. Racial differences in ITS have also been demonstrated.25 Finally, Perkins et al26 hypothesized that women’s smoking is less motivated by nicotine, suggesting sex differences in response to VLNCCs. Accordingly, we evaluated history of daily smoking, race/ethnicity, and sex as potential moderators of VLNCC effects.

Methods

This was a 12-week randomized, controlled, double-blind intervention trial that took place between June 2015 and July 2017. Analyses began August 2017. The trial protocol is available in Supplement 1. After a 2-week baseline period when participants smoked their own brand of cigarettes (provided free to parallel free cigarettes provided subsequently; this increased smoking27), participants were randomized (1:1, block size of 10, stratified by own-brand menthol preference) to receive VLNCCs or normal-nicotine-content cigarettes (NNCCs), blinded (identical in appearance), which they were to smoke exclusively for 10 weeks (eFigure 1 in Supplement 2). The study was approved by the University of Pittsburgh institutional review board. Participants provided written informed consent.

Participants

Adult ITS (defined as smoking any amount nondaily, 4-27 days per month) who were not planning to quit in the next 3 months were recruited through media and posters in Pittsburgh, Pennsylvania (see eTable 1 in Supplement 2 for detailed inclusion criteria). The study aimed to recruit 364 participants to achieve 80% power to detect a 20% treatment group difference in change in cigarettes per day (CPD) between baseline and weeks 9 to 10 postrandomization (eAppendix 13.3 in Supplement 1).

Procedures

At research site visits, participants were dispensed cigarettes, returned butts from smoked cigarettes, completed assessments, reported potential adverse events, provided urine specimens, and had carbon monoxide assessed in a breath sample. Participants received financial compensation for attending each visit for a total of $455 (eAppendix 11.0 in Supplement 1).

Cigarettes

Experimental cigarettes had either very low nicotine (VLNCCs: 0.07 mg nicotine delivery; NRC200, NRC201) or normal nicotine (NNCCs: 0.8 mg; NRC600, NRC601)28 and were identical in appearance; menthol smokers received menthol cigarettes. The tobacco’s nicotine content was manipulated genetically so participants could not extract normal amounts of nicotine from the VLNCCs.28 Cigarettes were provided by the National Institute of Drug Abuse and were approved for use as an Investigational Tobacco Product by the US Food and Drug Administration.

Measures

Cigarette consumption was the primary outcome and was assessed using 3 different methods: (1) timeline follow-back reports of the number of research and conventional cigarettes smoked each day since the prior visit (retrospective reports entered in calendar format)29-31; (2) counts of cigarette butts stored in plastic bags issued for each day; and (3) reports from participants each time they smoked via calls to an interactive voice response system; cell phones were provided as needed. Daily, participants received a call to report cigarettes not reported in real time. The 3 methods were highly concordant, yielding a highly reliable composite (mean) measure of daily consumption.32 Mean CPD was computed during 2-week blocks, relative to time of randomization.

Abstinence for the final 2-week block was also assessed, with biochemical validation via urinary total cotinine (<100 ng/mL; carbon monoxide, <8 ppm).27,33 (Data from ITS19 indicated that 14 days of abstinence was rare during ad libitum smoking.) In intent-to-treat (ITT) analyses, smoking was assumed for participants lost to observation.34,35 On exit from the study, participants were asked to rate their intention to quit smoking (1 indicating not at all and 5 indicating completely) in the succeeding month and in 6 months.

Cotinine concentrations (Total: Free + Glucuronide, ng/mL) in urine were assayed by liquid chromatography–mass spectrometry (Masonic Cancer Center, University of Minnesota)36,37 in spot urine samples at randomization and 2, 6, and 10 weeks later. We adapted for ITS published algorithms for imputing use of conventional cigarettes (cheating) in participants whose observed cotinine levels were implausible absent such behavior38,39 (eAppendix 2.4 in Supplement 1).

Participants were instructed not to use other nicotine products (ie, e-cigarettes, nicotine replacement products, smokeless tobacco, pipes, cigars, and hookah) but to report it if they did by timeline follow-back and interactive voice response. The higher reported value was analyzed. Adverse events, coded by the study physician blind to group assignment for seriousness and likely causality, were assessed by spontaneous report and abstracted from the Respiratory Health Questionnaire, as in Donny et al.11

Statistical Analysis

The primary analysis assessed change in mean CPD between baseline and the final 2 weeks (9-10) of treatment (selected as a direct, transparent measure of change), controlling for baseline CPD (because change may depend on the starting value). This ITT analysis used multiple imputation (25 iterations) of missing CPD data, using methods for monotone missingness (eAppendix 2.2 in Supplement 1).40 The imputation used multiple baseline variables to predict composite CPD for those participants with missing CPD values in weeks 9 to 10. Observed data included 3 participants who contributed more than 5 days of interactive voice response in weeks 9 to 10 despite not completing the study. The variables used in imputation were treatment group, age, education, sex, race/ethnicity, daily smoker history, income, number of years smoked, Nicotine Dependence Syndrome Scale score,41 Wisconsin Inventory of Smoking Dependence Motives Primary Dependence Motives score,42 exhaled breath carbon monoxide concentration, and prestudy mean and maximum CPD from a retrospective timeline follow-back assessment at enrollment. Sex, race/ethnicity, and history of daily smoking were analyzed as prespecified potential moderators of treatment effects in both ITT and complete-participants analyses.

Secondary analyses of cigarette consumption used linear mixed-effects models with random intercept and slope, with both linear and quadratic terms for 2-week intervals, to assess patterns of change in CPD throughout time.43 Treatment group differences in abstinence were modeled using logistic regression. Effect coding was used throughout.44

Sensitivity analyses addressed cheating with conventional cigarettes, both by self-report (analyses excluding conventional cigarettes or participants reporting ≥10% conventional cigarettes) and by inference from cotinine (eAppendix 2.3 in Supplement 1), as well as use of other nicotine-containing products. Differences in baseline smoking patterns were also examined as moderators, and analyses with log-transformed CPD data were also performed. Sensitivity analyses followed the baseline-to-end comparisons of the primary analysis using observed data.45 Statistical analyses used 2-sided tests at P < .05 (without multiplicity correction), using SAS, version 9.4 (SAS Institute, Inc).

Results

Between June 2015 and May 2017, 297 participants were enrolled, and 238 (80.1%) were randomized. The study did not reach its target sample size during the funding period, but the dropout rate was lower than anticipated. Randomization was successful in balancing the treatment groups, which were similar in demographics and smoking history (Table). The mean (SD) age of randomized participants was 37.9 (13.8) years, 120 (55%) were women, 61 (26%) were black, and 118 (50%) were converted ITS. Black race, heavier smoking, and lower education levels were overrepresented in participants who discontinued participation between enrollment and randomization (Table).

Figure 1 and eTable 2 in Supplement 2 show that loss to observation after randomization was similar in the 2 treatment arms. Participants with higher baseline CPD were more likely to be lost (OR, 1.12 per baseline CPD; 95% CI, 1.02-1.24; P = .02). Controlling for baseline CPD, sex, and history as a daily smoker were not associated with loss, but black individuals were more likely to be lost (22 [36.1%]) than white individuals (23 [15.1%]) (OR, 1.74; 95% CI, 1.34-5.55; P = .01). In a proportional hazards survival analysis controlling for baseline CPD, ongoing CPD during treatment as a time-varying covariate (essentially, response to treatment) did not predict subsequent loss to observation (adjusted hazard ratio for a 1-unit increase in CPD, 1.07; 95% CI, 0.95-1.21; P = .25).

Cigarette Consumption

In ITT analysis controlling for baseline CPD, the VLNCC group reduced consumption by 1.51 CPD more (95% CI, 0.86-2.17) than the NNCC group throughout 10 weeks (P < .001; Figure 2 and eTable 3 in Supplement 2). Mean change was −1.6 CPD with VLNCCs (95% CI, −1.1 to −2.0; −51% vs baseline) vs −0.05 with NNCCs (95% CI, −0.5 to 0.4; −2% vs baseline). Complete-case analysis showed substantively the same effect (Figure 2). The standardized effect size (Cohen d)47 was 0.6. Similar treatment effects were seen for log(CPD) (eFigure 2 in Supplement 2) and for both the proportion of days smoked (Figure 3) and cigarette consumption on smoking days (eFigure 3 in Supplement 2).

In planned subgroup analyses, VLNCC effects did not vary by sex or race/ethnicity (eFigures 4 and 5 in Supplement 2). Mean decrease among converted ITS was 1.89 CPD (95% CI, 0.80-2.97) more with VLNCC than NNCC, while mean decrease among native ITS without a history of daily smoking was 1.07 CPD (95% CI, 0.35-1.79). Converted ITS smoked more at baseline (as previously observed24) and accordingly had greater capacity to decrease smoking. When CPD was expressed as a percentage of baseline, converted ITS and native ITS showed very similar two-thirds decrease in CPD (eFigure 6 in Supplement 2). Similarly, heavier baseline smokers showed greater effects but not when expressed as a percentage of baseline, and those who smoked on more days showed a greater percentage of reductions in CPD (eFigures 7 and 8 in Supplement 2).

Analysis throughout 2-week periods also showed that CPD decreased significantly more in the VLNCC group over time (Figure 4). In the VLNCC group, cigarette consumption decreased steeply over the first 4 weeks of treatment (by 1.18 CPD [95% CI, 0.76-1.59] or 43.8% of baseline [95% CI, 23.3%-55.4%) and then leveled off (treatment × time interaction, linear P < .001, quadratic P = .002). A post hoc analysis showed no group differences in CPD change from week 5 on (β = .05; 95% CI, −0.04 to 0.13; P = .27).

Abstinence and Intention to Quit

Reported abstinence in the last 2 study weeks did not differ by treatment group in logistic regression models controlling for baseline CPD, either in ITT analysis counting lost participants as smoking (VLNCC, 12 of 118 [10.2%] and NNNC, 6 of 120 [5.0%]; OR, 1.78; 95% CI, 0.62-5.12; P = .28) or among completers (OR, 1.94; 95% CI, 0.67-5.63; P = .22). Intention to quit rating, assessed by t tests in 192 participants, mostly study completers, did not vary by treatment group either for the next month (VLNCC, 2.55 [95% CI, 2.27-2.84] and NNCC, 2.39 [95% CI, 2.12-2.66]; P = .41) or within 6 months (VLNCC, 2.86 [95% CI, 2.56-3.16] and NNCC, 2.73 [95% CI, 2.44-3.02]; P = .54).

Compliance

Overall, participants reported that 3.6% (95% CI, 2.4%-4.9%) of the cigarettes they smoked postrandomization were conventional cigarettes. This cheating was concentrated in 27 participants (19 in the VLNCC group and 8 in the NNCC group) who reported 10% or more of their smoking was nonresearch cigarettes; they accounted for 69% of all nonresearch cigarettes. The odds of VLNCC participants’ being in this group were 2.69 (95% CI, 1.13-6.41) times greater than for NNCC participants (P = .03). Primary treatment effects were not substantively changed by excluding these participants or by excluding conventional cigarettes from CPD calculations (eFigures 9 and 10 in Supplement 2).

More VLNCC (38 of 107 [35.5%]) than NNCC participants (17 of 108 [15.7%]) were inferred to have cheated based on at least 1 of 3 postrandomization cotinine values (OR, 2.95; 95% CI, 1.54-5.66). As a sensitivity test, we reassessed change in CPD based on the noncheating participants; the effect was still significant (eFigure 11 in Supplement 2) (95% CI, 0.53-2.05; P < .01). Urinary cotinine concentrations were also analyzed as continuous variables and showed significantly greater decreases in the VLNCC group (eFigures 12 and 13 in Supplement 2).

The most commonly used nicotine product was e-cigarettes on 0.8% of study days (148 of 17 738). Among participants not using e-cigarettes at enrollment or baseline, new use of e-cigarettes was statistically significantly more common in the VLNCC group, particularly among heavier smokers (eFigure 14 in Supplement 2). Excluding participants who increased postrandomization use of other tobacco/nicotine products by 1 or more unit per week did not substantively change treatment effect estimates (eFigure 15 in Supplement 2).

Subjective Discernment of Condition

At exit, participants were asked to guess which cigarettes they had been randomly assigned. Participants in the VLNCC group were more likely to guess correctly (χ2 test P < .01; 73 [80.2%] vs 39 [39.0%]; OR, 6.34; 95% CI, 3.30-12.20).

Adverse Events

Overall, 83 randomized participants (34.9%) reported an adverse event, of which 34.8% (n = 41) were in the VLNCC group and 35.0% (n = 42) were in the NNCC group (χ2 test P = .97). Three (1 in VLNCC) reported serious adverse events, none related to study treatment.

Discussion

This study demonstrates that switching nondaily smokers to VLNCCs leads to a reduction in cigarette consumption, consistent with the role of nicotine as a primary reinforcer for smoking. Smokers switched to VLNCCs reduced their smoking by 51%, a large effect.47 Not only did ITS reduce the number of cigarettes they consumed, they also reduced the number of days on which they smoked, suggesting substantial diminution in reinforcement. The VLNCC effect was most prominent in ITS who smoked on more days at baseline and thus were closer to daily smoking. These data extend to nondaily smokers the findings of Donny et al11 and others9,10,46 who demonstrated reductions in cigarette consumption (although smaller on a percentage basis) in daily smokers, who may experience withdrawal when nicotine-deprived. This is important because nondaily smokers now constitute 25% to 33% of adult smokers in the United States,12-14 making it crucial that effects on this fraction of the smoker population be considered in tobacco policy.

Our results suggest that a policy mandating VLNCCs might reduce consumption without leading to quitting in ITS. Reductions in CPD after switching to VLNCCs occurred relatively quickly over 4 weeks, then levelled off, not suggesting a trajectory that would lead to smoking no cigarettes. Despite a trend for more abstinence in the VLNCC group, most VLNCC participants continued to smoke and did not have greater intention to quit than NNCC participants after the 10-week intervention. A similar pattern of reduced consumption without extinction was observed in daily smokers using VLNCCs,11 suggesting that VLNCCs maintain enough reinforcement to sustain smoking for some individuals for some time or that occasional smoking of conventional cigarettes (which would be difficult if a low-nicotine policy were implemented) maintained smoking. Adverse events were uncommon and did not differ by treatment. Our results mitigate concern that ITS might increase smoking when switched to VLNCCs. That ITS respond to VLNCCs in a manner similar to daily smokers suggests that a universal policy is possible.

Although treatment assignment was blinded, ITS smoking VLNCCs were often able to guess they were in the low-nicotine group. This is not surprising, as nicotine’s sensory and psychoactive effects are discernable.41

Although most participants reported smoking only the experimental cigarettes, some admitted cheating by smoking conventional cigarettes, and others were assumed to have cheated based on their cotinine levels. Importantly, VLNCCs caused reductions in CPD even among those who did not get nicotine elsewhere. Also of interest, cheating was more common in the VLNCC group, as was new use of e-cigarettes, suggesting these behaviors were motivated by nicotine-seeking.

This study contributes to understanding ITS smoking behavior. Studies have documented that ITS inhale, absorb normal amounts of nicotine per cigarette, and metabolize it normally18 but do not experience withdrawal.19 We had hypothesized that ITS smoke for the acute reinforcing effects of nicotine,42 while others have suggested that social or sensory factors might be primary motivators.48,49 In this study, ITS smoking VLNCCs reduced their smoking and were more likely to seek nicotine elsewhere, suggesting that nicotine-seeking motivates smoking even absent of nicotine dependence. We have previously suggested that ITS might seek immediate reinforcement from acute nicotine administration in particular settings, with this stimulus-driven nicotine-seeking constituting a form of dependence.50

Switching to VLNCCs seemingly motivated ITS to seek nicotine elsewhere. Both the original proposal for a low-nicotine cigarette policy4 and the US Food and Drug Administration’s policy announcement7,8 contemplate an integrated policy of reducing nicotine in cigarettes while making alternative nicotine products available51,52 to help smokers move away from cigarettes, as smoking carries far greater risk than nicotine itself.53 Indeed, we saw movement toward e-cigarettes among heavier nondaily smokers who received VLNCCs.

Limitations and Strengths

The study had limitations. Participants were from a single geographic region and were not nationally representative. For example, Hispanic and higher-income ITS were both underrepresented, and black ITS were overrepresented (C. M. Reyes-Guzman, PhD, written communication, March 2018), even after some differential dropout among black participants. Differential dropout of black participants and heavier smokers before randomization underscores that the randomized cohort is not a random sample. Frequent users of other nicotine products were excluded, also limiting generalizability with regard to how VLNCCs might influence a shift toward use of such products. The enrollment target was not met; however, the sample size was adequate to detect significant treatment effects on CPD. The switch to VLNCCs was not always complete: some participants admitted cheating by smoking conventional cigarettes, and urinary cotinine data suggested additional cheating. However, VLNCC effects were seen even among compliant participants.

A strength of the study was the randomized double-blind design, which allowed the observed changes to be attributed to the cigarettes’ nicotine levels. Also, the provision of free cigarettes was kept constant between baseline and treatment phases11; thus, the change after randomization could be attributed completely to the experimental manipulation without being complicated by increased consumption when cigarettes were free.27 Another strength was the inclusion of 3 different measures of cigarette consumption, which demonstrated marked agreement.32 Greater CPD reductions in the VLNCC group compared with NNCC controls were apparent across key subgroups defined by race/ethnicity, sex, and prior daily smoking history.

Conclusions

This randomized clinical trial was conducted in nondaily smokers, a substantial fraction of the smoking population whose response to VLNCCs had not previously been studied. Switching to VLNCCs resulted in reduced cigarette consumption among these nondaily smokers, much as it does in daily smokers,11 and there was no evidence of adverse effects such as increased smoking. That daily and nondaily smokers react similarly suggests that reduced nicotine policy need not be concerned about differential response among nondaily smokers. However, while the stated goal of such policy is to increase quitting,7,8,54 the observed reductions in smoking leveled off after several weeks and switching to VLNCCs did not significantly increase quitting, suggesting that the data do not yet support that expected policy outcome.

Back to top
Article Information

Accepted for Publication: May 24, 2018.

Corresponding Author: Saul Shiffman, PhD, 130 N Bellefield Ave, Ste 510, Pittsburgh, PA 15213 (shiffman@pinneyassociates.com).

Published Online: June 14, 2018. doi:10.1001/jamapsychiatry.2018.1831

Author Contributions: Drs Shiffman and Kurland had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Shiffman, Kurland.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Shiffman, Kurland, Scholl.

Critical revision of the manuscript for important intellectual content: Shiffman, Kurland.

Statistical analysis: Shiffman, Kurland, Mao.

Obtained funding: Shiffman.

Administrative, technical, or material support: Scholl.

Supervision: Scholl.

Conflict of Interest Disclosures: Dr Shiffman, through Pinney Associates, consults on tobacco cessation and harm reduction (including nicotine replacement therapy and digital vapor products; by contract, combusted cigarettes are excluded) to Niconovum USA, R. J. Reynolds Vapor Company, and RAI Services Company, all subsidiaries of Reynolds American Inc and British American Tobacco. Previously, Dr Shiffman consulted to NJOY on e-cigarettes and to GlaxoSmithKline Consumer Healthcare on smoking cessation medications and treatments. Dr Shiffman holds patents for a novel nicotine smoking cessation medication that is not under commercial development. No other disclosures are reported.

Funding/Support: This work was supported by a grant (Dr Shiffman) from the National Cancer Institute at the National Institutes of Health and the Center for Tobacco Products at the Food and Drug Administration, awarded as a supplement to grant P30CA047904. The grant supported use of the University of Pittsburgh Medical Center Hillman Cancer Center Biostatistics Shared Resource Facility.

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the US Food and Drug Administration.

Additional Contributions: We thank Allison Brown, MA, for assistance overseeing the study; David Colarusso, BS, Corinne Hogge, BS, and Ian Jutsum, BA, research assistants who conducted research sessions; James Moorehead, BS, for data management and preparation; Alexsys Hoesch, BA, for administrative assistance, and Esa Davis, MD, for medical oversight. The above named contributors are all at the University of Pittsburgh and were compensated by the sponsor in accordance with the grant. We also appreciate the contributions of members of the study Data and Safety Monitoring Board, Peter Callas, PhD (University of Vermont), Matthew Carpenter, PhD (Medical University of South Carolina), Jonathan Foulds, PhD (Pennsylvania State University), and John Hughes, MD (University of Vermont), who also reviewed and commented on a draft of this article. These individuals were not compensated.

References
1.
Jacobs  EJ, Newton  CC, Carter  BD,  et al.  What proportion of cancer deaths in the contemporary United States is attributable to cigarette smoking?  Ann Epidemiol. 2015;25(3):179-182.e1.PubMedGoogle ScholarCrossref
2.
Benowitz  NL.  Nicotine addiction.  N Engl J Med. 2010;362(24):2295-2303. doi:10.1056/NEJMra0809890PubMedGoogle ScholarCrossref
3.
Stolerman  IP, Jarvis  MJ.  The scientific case that nicotine is addictive.  Psychopharmacology (Berl). 1995;117(1):2-10. doi:10.1007/BF02245088PubMedGoogle ScholarCrossref
4.
Benowitz  NL, Henningfield  JE.  Establishing a nicotine threshold for addiction: the implications for tobacco regulation.  N Engl J Med. 1994;331(2):123-125. doi:10.1056/NEJM199407143310212PubMedGoogle ScholarCrossref
5.
Henningfield  JE, Benowitz  NL, Slade  J, Houston  TP, Davis  RM, Deitchman  SD.  Reducing the addictiveness of cigarettes: Council on Scientific Affairs, American Medical Association.  Tob Control. 1998;7(3):281-293. doi:10.1136/tc.7.3.281PubMedGoogle ScholarCrossref
6.
Family Smoking Prevention and Tobacco Control Act and Federal Retirement Reform, HR 1256, 111th Cong, 1st Sess (PA 1995).
7.
Tobacco product standard for nicotine level of combusted cigarettes. Fed Regist. 2018;83:11818-11843. To be codified at 21 CFR §1130.
8.
FDA announces comprehensive regulatory plan to shift trajectory of tobacco-related disease, death [press release]. Silver Spring, MD: US Food and Drug Administration; July 28, 2017.
9.
Benowitz  NL, Hall  SM, Stewart  S, Wilson  M, Dempsey  D, Jacob  P  III.  Nicotine and carcinogen exposure with smoking of progressively reduced nicotine content cigarette.  Cancer Epidemiol Biomarkers Prev. 2007;16(11):2479-2485. doi:10.1158/1055-9965.EPI-07-0393PubMedGoogle ScholarCrossref
10.
Donny  EC, Houtsmuller  E, Stitzer  ML.  Smoking in the absence of nicotine: behavioral, subjective and physiological effects over 11 days.  Addiction. 2007;102(2):324-334. doi:10.1111/j.1360-0443.2006.01670.xPubMedGoogle ScholarCrossref
11.
Donny  EC, Denlinger  RL, Tidey  JW,  et al.  Randomized trial of reduced-nicotine standards for cigarettes.  N Engl J Med. 2015;373(14):1340-1349. doi:10.1056/NEJMsa1502403PubMedGoogle ScholarCrossref
12.
Jamal  A, King  BA, Neff  LJ, Whitmill  J, Babb  SD, Graffunder  CM.  Current cigarette smoking among adults: United States, 2005-2015.  MMWR Morb Mortal Wkly Rep. 2016;65(44):1205-1211. doi:10.15585/mmwr.mm6544a2PubMedGoogle ScholarCrossref
13.
US Department of Health and Human Services.  Results from the 2013 national survey on drug use and health: summary of national findings. https://www.samhsa.gov/data/sites/default/files/NSDUHresultsPDFWHTML2013/Web/NSDUHresults2013.pdf. Accessed June 5, 2018.
14.
Reyes-Guzman  CM, Pfeiffer  RM, Lubin  J,  et al.  Determinants of light and intermittent smoking in the United States: results from three pooled national health surveys.  Cancer Epidemiol Biomarkers Prev. 2017;26(2):228-239. doi:10.1158/1055-9965.EPI-16-0028PubMedGoogle ScholarCrossref
15.
Inoue-Choi  M, Liao  LM, Reyes-Guzman  C, Hartge  P, Caporaso  N, Freedman  ND.  Association of long-term, low-intensity smoking with all-cause and cause-specific mortality in the National Institutes of Health-AARP Diet and Health Study.  JAMA Intern Med. 2017;177(1):87-95. doi:10.1001/jamainternmed.2016.7511PubMedGoogle ScholarCrossref
16.
Schane  RE, Ling  PM, Glantz  SA.  Health effects of light and intermittent smoking: a review.  Circulation. 2010;121(13):1518-1522. doi:10.1161/CIRCULATIONAHA.109.904235PubMedGoogle ScholarCrossref
17.
Tindle  HA, Shiffman  S.  Smoking cessation behavior among intermittent smokers versus daily smokers.  Am J Public Health. 2011;101(7):e1-e3. doi:10.2105/AJPH.2011.300186PubMedGoogle ScholarCrossref
18.
Shiffman  S, Dunbar  MS, Benowitz  NL.  A comparison of nicotine biomarkers and smoking patterns in daily and nondaily smokers.  Cancer Epidemiol Biomarkers Prev. 2014;23(7):1264-1272. doi:10.1158/1055-9965.EPI-13-1014PubMedGoogle ScholarCrossref
19.
Shiffman  S, Dunbar  MS, Tindle  HA, Ferguson  SG.  Nondaily smokers’ experience of craving on days they do not smoke.  J Abnorm Psychol. 2015;124(3):648-659. doi:10.1037/abn0000063PubMedGoogle ScholarCrossref
20.
Shiffman  S, Paty  JA, Gnys  M, Kassel  JD, Elash  C.  Nicotine withdrawal in chippers and regular smokers: subjective and cognitive effects.  Health Psychol. 1995;14(4):301-309. doi:10.1037/0278-6133.14.4.301PubMedGoogle ScholarCrossref
21.
Russell  MA.  Cigarette smoking: natural history of a dependence disorder.  Br J Med Psychol. 1971;44(1):1-16. doi:10.1111/j.2044-8341.1971.tb02141.xPubMedGoogle ScholarCrossref
22.
Shiffman  S, Dunbar  MS, Li  X,  et al.  Smoking patterns and stimulus control in intermittent and daily smokers.  PLoS One. 2014;9(3):e89911. doi:10.1371/journal.pone.0089911PubMedGoogle ScholarCrossref
23.
Shiffman  S, Tindle  H, Li  X, Scholl  S, Dunbar  M, Mitchell-Miland  C.  Characteristics and smoking patterns of intermittent smokers.  Exp Clin Psychopharmacol. 2012;20(4):264-277. doi:10.1037/a0027546PubMedGoogle ScholarCrossref
24.
Shiffman  S, Dunbar  MS, Kirchner  TR,  et al.  Cue reactivity in converted and native intermittent smokers.  Nicotine Tob Res. 2015;17(1):119-123. doi:10.1093/ntr/ntu147PubMedGoogle ScholarCrossref
25.
Trinidad  DR, Pérez-Stable  EJ, Emery  SL, White  MM, Grana  RA, Messer  KS.  Intermittent and light daily smoking across racial/ethnic groups in the United States.  Nicotine Tob Res. 2009;11(2):203-210. doi:10.1093/ntr/ntn018PubMedGoogle ScholarCrossref
26.
Perkins  KA, Jacobs  L, Sanders  M, Caggiula  AR.  Sex differences in the subjective and reinforcing effects of cigarette nicotine dose.  Psychopharmacology (Berl). 2002;163(2):194-201. doi:10.1007/s00213-002-1168-1PubMedGoogle ScholarCrossref
27.
Shiffman  S, Scholl  S.  Increases in cigarette consumption and decreases in smoking intensity when non-daily smokers are provided with free cigarettes.  Nicotine Tob Res. 2017; Advance online publication. doi:10.1093/ntr/ntx221PubMedGoogle Scholar
28.
Twenty Second Century. SPECTRUM Government Research Cigarettes. http://www.xxiicentury.com/.
29.
Blank  MD, Breland  AB, Enlow  PT, Duncan  C, Metzger  A, Cobb  CO.  Measurement of smoking behavior: comparison of self-reports, returned cigarette butts, and toxicant levels.  Exp Clin Psychopharmacol. 2016;24(5):348-355. doi:10.1037/pha0000083PubMedGoogle ScholarCrossref
30.
Sobell  LC, Sobell  MB, Maisto  SA. Time-line follow-back assessment methods. In: Lettieri  DJ, Sayers  MA, Nelson  JE, eds.  NIAA Treatment Handbook Series: Vol. 2, Alcoholism Treatment Assessment Research Instruments. Washington, DC: National Institute on Alcoholism and Drug Abuse DHHS Publications; 1985:530-534.
31.
Wray  JM, Gass  JC, Miller  EI, Wilkins  DG, Rollins  DE, Tiffany  ST.  A comparative evaluation of self-report and biological measures of cigarette use in nondaily smokers.  Psychol Assess. 2016;28(9):1043-1050. doi:10.1037/pas0000227PubMedGoogle ScholarCrossref
32.
Shiffman  S, Scholl  SM.  Three approaches to quantifying cigarette consumption: data from nondaily smokers.  Psychol Addict Behav. 2018;32(2):249-254. doi:10.1037/adb0000346PubMedGoogle ScholarCrossref
33.
SRNT Subcommittee on Biochemical Verification.  Biochemical verification of tobacco use and cessation.  Nicotine Tob Res. 2002;4(2):149-159. doi:10.1080/14622200210123581PubMedGoogle ScholarCrossref
34.
Hughes  JR, Keely  JP, Niaura  RS, Ossip-Klein  DJ, Richmond  RL, Swan  GE.  Measures of abstinence in clinical trials: issues and recommendations.  Nicotine Tob Res. 2003;5(1):13-25. doi:10.1080/1462220031000070552PubMedGoogle ScholarCrossref
35.
West  R, Hajek  P, Stead  L, Stapleton  J.  Outcome criteria in smoking cessation trials: proposal for a common standard.  Addiction. 2005;100(3):299-303. doi:10.1111/j.1360-0443.2004.00995.xPubMedGoogle ScholarCrossref
36.
Murphy  SE, Park  SS, Thompson  EF,  et al.  Nicotine N-glucuronidation relative to N-oxidation and C-oxidation and UGT2B10 genotype in five ethnic/racial groups.  Carcinogenesis. 2014;35(11):2526-2533. doi:10.1093/carcin/bgu191PubMedGoogle ScholarCrossref
37.
Pitt  JJ.  Principles and applications of liquid chromatography-mass spectrometry in clinical biochemistry.  Clin Biochem Rev. 2009;30(1):19-34.PubMedGoogle Scholar
38.
Nardone  N, Donny  EC, Hatsukami  DK,  et al.  Estimations and predictors of non-compliance in switchers to reduced nicotine content cigarettes.  Addiction. 2016;111(12):2208-2216. doi:10.1111/add.13519PubMedGoogle ScholarCrossref
39.
Benowitz  NL, Nardone  N, Hatsukami  DK, Donny  EC.  Biochemical estimation of noncompliance with smoking of very low nicotine content cigarettes.  Cancer Epidemiol Biomarkers Prev. 2015;24(2):331-335. doi:10.1158/1055-9965.EPI-14-1040PubMedGoogle ScholarCrossref
40.
Wang  B, Fang  Y, Jin  M.  Combining type-III analyses from multiple imputations. http://support.sas.com/resources/papers/proceedings14/1543-2014.pdf. Accessed June 5, 2018.
41.
Hatsukami  DK, Heishman  SJ, Vogel  RI,  et al.  Dose-response effects of spectrum research cigarettes.  Nicotine Tob Res. 2013;15(6):1113-1121. doi:10.1093/ntr/nts247PubMedGoogle ScholarCrossref
42.
Shiffman  S, Dunbar  MS, Scholl  SM, Tindle  HA.  Smoking motives of daily and non-daily smokers: a profile analysis.  Drug Alcohol Depend. 2012;126(3):362-368. doi:10.1016/j.drugalcdep.2012.05.037PubMedGoogle ScholarCrossref
43.
Singer  JM, Rocha  FMM, Nobre  JS.  Graphical tools for detecting departures from linear mixed model assumptions and some remedial measures.  Int Stat Rev. 2017;85(2):290-324. doi:10.1111/insr.12178Google ScholarCrossref
44.
Edwards  AL.  Multiple Regression and the Analysis of Variance and Covariance New York. NY: Freeman; 1985.
45.
 Heavy: Robust Estimation Using Heavy-Tailed Distributions. CRAN. https://cran.r-project.org/web/packages/heavy/index.html. Accessed June 5, 2018,
46.
Hatsukami  DK, Kotlyar  M, Hertsgaard  LA,  et al.  Reduced nicotine content cigarettes: effects on toxicant exposure, dependence and cessation.  Addiction. 2010;105(2):343-355. doi:10.1111/j.1360-0443.2009.02780.xPubMedGoogle ScholarCrossref
47.
Cohen  J.  Statistical power analysis for the behavioral sciences. Hillsdale, NJ: Lawrence Erlbaum Associates; 1987.
48.
Rose  JE.  Nicotine and nonnicotine factors in cigarette addiction.  Psychopharmacology (Berl). 2006;184(3-4):274-285. doi:10.1007/s00213-005-0250-xPubMedGoogle ScholarCrossref
49.
Schane  RE, Glantz  SA, Ling  PM.  Nondaily and social smoking: an increasingly prevalent pattern.  Arch Intern Med. 2009;169(19):1742-1744. doi:10.1001/archinternmed.2009.315PubMedGoogle ScholarCrossref
50.
Shiffman  S, Dunbar  MS, Ferguson  SG.  Stimulus control in intermittent and daily smokers.  Psychol Addict Behav. 2015;29(4):847-855. doi:10.1037/adb0000052PubMedGoogle ScholarCrossref
51.
Farsalinos  KE, Polosa  R.  Safety evaluation and risk assessment of electronic cigarettes as tobacco cigarette substitutes: a systematic review.  Ther Adv Drug Saf. 2014;5(2):67-86. doi:10.1177/2042098614524430PubMedGoogle ScholarCrossref
52.
Polosa  R, Rodu  B, Caponnetto  P, Maglia  M, Raciti  C.  A fresh look at tobacco harm reduction: the case for the electronic cigarette.  Harm Reduct J. 2013;10:19. doi:10.1186/1477-7517-10-19PubMedGoogle ScholarCrossref
53.
Royal College of Physicians.  Nicotine without smoke: tobacco harm reduction. https://www.rcplondon.ac.uk/projects/outputs/nicotine-without-smoke-tobacco-harm-reduction-0. Accessed June 5, 2018.
54.
Apelberg  BJ, Feirman  SP, Salazar  E,  et al.  Potential public health effects of reducing nicotine levels in cigarettes in the United States.  N Engl J Med. 2018;378(18):1725-1733. doi:10.1056/NEJMsr1714617PubMedGoogle ScholarCrossref
×